Main Report (5.722Mb)

(1)

ICES Advisory Committee on Fishery Management ICES CM 2005/ACFM:17 R

^EF

. I

Report of the Working Group on North Atlantic Salmon (WGNAS)

5-14 April 2005

Nuuk, Greenland

(2)

Conseil International pour l’Exploration de la Mer

H.C. Andersens Boulevard 44-46 DK-1553 Copenhagen V Denmark

Telephone (+45) 33 38 67 00 Telefax (+45) 33 93 42 15 www.ices.dk

[email protected]

Recommended format for purposes of citation:

ICES. 2005. Report of the Working Group on North Atlantic Salmon (WGNAS), 5-14 April 2005, Nuuk, Greenland. ICES CM 2005/ACFM:17. 291pp.

For permission to reproduce material from this publication, please apply to the General Secre- tary.

The document is a report of an Expert Group under the auspices of the International Council for the Exploration of the Sea and does not necessarily represent the views of the Council.

(3)

1 INTRODUCTION

1.1 Main Tasks

At its 2004 Statutory Meeting, ICES resolved (C. Res. 2004/2ACFM04) that the Working Group on North Atlantic Salmon [WGNAS] (Chair: Dr W Crozier, UK) will meet in Nuuk, Greenland, from the 4-14th April 2005 to consider questions posed to ICES by the North At- lantic Salmon Conservation Organisation (NASCO). The terms of reference and sections of the report in which the answers are provided, follow:

a) With respect to Atlantic salmon in the North Atlantic area: Section 2 i. provide an overview of salmon catches and landings, including unreported catches by

country and catch and release, and worldwide production of farmed and ranched salmon in 2004;

2.1, 2.2

ii. report on significant developments which might assist NASCO with the management of salmon stocks;

2.4

iii. provide a compilation of tag releases by country in 2004; 2.7 iv. identify relevant data deficiencies, monitoring needs and research requirements. 6

b) With respect to Atlantic salmon in the North-East Atlantic Commission area: Section 3 i. describe the key events of the 2004 fisheries and the status of the stocks; 3.9 ii. provide any new information on the extent to which the objectives of any significant man-

agement measures introduced in recent years have been achieved;

3.10

.iii. further develop the age-specific stock conservation limits where possible based upon individual river stocks;

3.3

.iv. provide catch options or alternative management advice, if possible based on forecasts of PFA, for northern and southern stocks, with an assessment of risks relative to the objective of exceeding stock conservation limits and advise on the implications of these options for stock rebuilding;

3.4, 3.6

v. provide an estimate of by-catch of salmon in pelagic fisheries. 3.11

c) With respect to Atlantic salmon in the North American Commission area: Section 4 i. describe the key events of the 2004 fisheries and the status of the stocks; 4.9 ii. provide any new information on the extent to which the objectives of any significant man-

4.10

iii. update age-specific stock conservation limits based on new information as available; 4.3 iv. provide catch options or alternative management advice with an assessment of risks rela-

tive to the objective of exceeding stock conservation limits and advise on the implications of these options ff for stock rebuilding;

4.4 and 4.6

v. provide an analysis of any new biological and/or tag return data, to identify the origin and biological characteristics of Atlantic salmon caught at St Pierre and Miquelon.

4.11

(8)

d) With respect to Atlantic salmon in the West Greenland Commission area: Section 5 i. describe the events of the 2004 fisheries and the status of the stocks; 5.9 ii. provide any new information on the extent to which the objectives of any significant man-

5.12

iii. provide information on the origin of Atlantic salmon caught at West Greenland at a finer resolution than continent of origin (river stocks, country or stock complexes);

5.9

iv. provide catch options or alternative management advice with an assessment of risks relative to the objective of exceeding stock conservation limits and advise on the implications of these options for stock rebuilding.

5.4 and 5.6

Notes:

NASCO’s International Atlantic Salmon research Board’s inventory of on-going re- search relating to salmon mortality in the sea will be provided to ICES to assist it in this task.

In the responses to questions b.i, c.i and d.i ICES is asked to provide details of catch, gear, effort, composition and origin of the catch and rates of exploitation. For homewater fisheries, the information provided should indicate the location of the catch in the following categories: in-river; estuarine; and coastal. Any new information on non-catch fishing mortality of the salmon gear used and on the bycatch of other species in salmon gear, and of salmon in any existing and new fisheries for other species is also requested.

In response to questions b.iv, c.iv and d.iv provide a detailed explanation and critical examination of any changes to the models used to provide catch advice.

In response to question d.i, ICES is requested to provide a brief summary of the status of North American and North-East Atlantic salmon stocks. The detailed information on the status of these stocks should be provided in response to questions b.i and c.i.

The Working Group considered 41 Working Documents submitted by participants (Appen- dix 1); other references cited in the report are given in Appendix 2. A full address list for the participants is provided in Appendix 3.

1.2 Participants

Crozier, W (Chair) UK (Northern Ireland) Chaput, G. Canada

Erkinaro, J. Finland Gudbergsson, G. Iceland Hansen, L.P. Norway Siegstad, H. Greenland Carl, J. Greenland Holm, M. Norway Legault, C. USA

MacLean, J.C. UK (Scotland) Meerburg, D.J. Canada Ó Maoiléidigh, N. Ireland Prusov, S. Russia Reddin, D.G. Canada

Russell, I.C. UK (England & Wales)

(9)

Sheehan, T. USA

Smith, G.W. UK (Scotland) Trial, J. USA

Vauclin, V. France

1.3 Management framework for salmon in the North Atlantic

The advice generated by ICES is in response to terms of reference posed by the North Atlantic Salmon Conservation Organisation (NASCO), pursuant to its role in international management of salmon. NASCO was set up in 1984 by international convention (the Convention for the Conservation of Salmon in the North Atlantic Ocean), with a responsibility for the conservation, restoration, enhancement, and rational management of wild salmon in the North Atlantic. While sovereign states retain their role in the regulation of salmon fisheries for salmon originating from their own rivers, distant water salmon fisheries, such as those at Greenland and Faroes, which take salmon originating from rivers of another Party are regulated by NASCO under the terms of the Convention. NASCO now has seven Parties that are signatories to the Convention, including the EU which represents its Member States.

NASCO discharges these responsibilities via three Commission areas shown below:

1.4 Management objectives

NASCO (NASCO CNL31.210) has identified the primary management objective of that organisation as:

“To contribute through consultation and co-operation to the conservation, restoration, enhancement and rational management of salmon stocks taking into account the best scientific advice available”.

NASCO further stated that “the Agreement on the Adoption of a Precautionary Approach states that an objective for the management of salmon fisheries is to provide the diversity and abundance of salmon stocks” and NASCOs Standing Committee on the Precautionary Approach interpreted this as being “to maintain both the productive capacity and diversity of salmon stocks”.

(10)

NASCO’s Action Plan for Application of the Precautionary Approach (NASCO, 1999) provides interpretation of how this is to be achieved, as follows:

• “Management measures should be aimed at maintaining all stocks above their conservation limits by the use of management targets”.

• Socio-economic factors could be taken into account in applying the Precautionary Approach to fisheries management issues”:

• “The precautionary approach is an integrated approach that requires, inter alia, that stock rebuilding programmes (including as appropriate, habitat improvements, stock enhancement, and fishery management actions) be developed for stocks that are below conservation limits”.

1.5 Reference points and application of precaution

Conservation limits (CLs) for North Atlantic salmon stock complexes have been defined by ICES as the level of stock (number of spawners) that will achieve long term average maximum sustainable yield (MSY), as derived from the adult to adult stock and recruitment relationship (Ricker, 1975; ICES, 1993). NASCO has adopted this definition of CLs (NASCO, 1998). Therefore, the CL is a limit reference point (Slim) which should be avoided with high probability. Management advice for Atlantic salmon is referenced to the Slim conservation limit, therefore stocks assessed here are reported as being outside precautionary limits when the confidence limits of the most recent stock estimate includes Slim.

Management targets have not yet been defined for North Atlantic salmon stocks. When these have been defined they will play an important role in ICES advice.

For the assessment of the status of stocks and advice on management of national components and geographical groupings of the stock complexes in the NEAC area, where there are no specific management objectives:

• ICES requires that the lower bound of the 95% confidence interval of the current estimate of spawners is above the CL for the stock to be considered at full reproductive capacity.

• When the lower bound of the confidence limit is below the CL, but the mid point is above, then ICES considers the stock to be at risk of suffering reduced reproductive capacity.

• Finally, when the mid point is below the CL, ICES considers the stock to suffer reduced reproductive capacity.

It should be noted that this is equivalent to the ICES precautionary target reference points (Spa). Therefore, stocks are regarded by ICES as being at full reproductive capacity only if they are above the precautionary reference point (Spa). This approach parallels the use of precautionary reference points used for the provision of catch advice for other fish stocks in the ICES area.

For catch advice on fish exploited at West Greenland (non maturing 1SW fish from North America and non maturing 1SW fish from Southern NEAC), ICES has adopted, a risk level of 75% (ICES, 2003) as part of an agreed management plan. ICES applies the same level of risk aversion for catch advice for homewater fisheries on the North American stock complex.

(11)

2 ATLANTIC SALMON IN THE NORTH ATLANTIC AREA

2.1

Catches of North Atlantic salmon

2.1.1

Nominal catches of salmon

The nominal catch of a fishery is defined as the round, fresh weight of fish that are caught and retained. Total nominal catches of salmon reported by country in all fisheries for 1960–2004 are given in Table 2.1.1.1. Catch statistics in the North Atlantic also include fish farm escapees and, in some north-east Atlantic countries, relatively small numbers of ranched fish (see Section 2.2.2). Catch and release has become increasingly commonplace in some countries, but these fish do not appear in the nominal catches (see Section 2.1.2).

Icelandic catches have traditionally been split into two separate categories, wild and ranched, reflecting the fact that Iceland has been the only North Atlantic country where large-scale ranching has been undertaken with the specific intention of harvesting all returns at the release site. However, the release of smolts for ranching purposes ceased in Iceland in 1998. While ranching does occur in some other countries, this is on a much smaller scale. Some of these operations are experimental and at others harvesting does not occur solely at the release site.

The ranched component in these countries has therefore been included in the nominal catch.

Figure 2.1.1.1 shows the nominal catch data grouped by the following areas: ‘Northern Europe’ (Norway, Russia, Finland, Iceland, Sweden and Denmark); ‘Southern Europe’ (Ire- land, UK (Scotland), UK (England and Wales), UK (Northern Ireland), France and Spain);

‘North America’ (including Canada, USA and St Pierre et Miquelon (France)); and

‘Greenland and Faroes’.

The provisional total nominal catch for 2004 was 2099 t, 357 t below the confirmed catch for 2003 (2456 t) and the lowest in the time series. The 2004 catch was over 550 t below the average of the last five years (2664 t), and almost 800 t below the average of the last 10 years (2881 t). Nominal catches were below the previous five- and ten-year averages in most countries, and were the lowest recorded in the time series in three countries.

Nominal catches in homewater fisheries split, where available, by sea-age or size category are presented in Table 2.1.1.2 (weight only). The data for 2004 are provisional and, as in Table 2.1.1.1, include both wild and reared salmon and fish farm escapees in some countries. A more detailed breakdown, providing both numbers and weight for different sea-age groups for most countries, is provided at Appendix 4. Countries use different methods to partition their catches by sea-age class and these are outlined in the footnotes to Appendix 4. The composition of catches in different areas is discussed in more detail in Sections 3, 4, and 5.

Table 2.1.1.3 presents the nominal catch by country in homewater fisheries partitioned accord- ing to whether the catch was taken in coastal, estuarine or riverine areas. Overall, coastal fisheries accounted for 50% of catches in North East Atlantic countries in 2004, in-river fisheries 42% and estuarine fisheries 7%. In North America, coastal fisheries accounted for 17% of the catch in 2004, while in-river fisheries took 66% and estuarine fisheries 18%.

There is considerable variability in the percentage of the catch taken in different fisheries between individual countries. For some countries the entire catch is taken in freshwater, while in other countries the majority of the catch is taken in coastal waters (Figure 2.1.1.2). Data ag- gregated by region are presented in Figure 2.1.1.3 for the period 1995–2004. Overall, in the NEAC northern area (Iceland, Norway, Russia, Finland and Sweden) around half the catch has been taken in coastal waters and half in rivers, although there are no coastal catches in Iceland and Finland. Estuarine catches have comprised no more than 2% of the total in this area. In the

(12)

NEAC southern area (France, Ireland, Spain, UK (N. Ireland), UK (Scotland) and UK (Eng- land & Wales)) most fish (50–64%) have been taken in coastal fisheries with riverine fisheries comprising around 30% and estuarine fisheries under 20%. In North America, the majority of the catch has been taken in freshwater (66–to 77 % over the period 1999–2004).

2.1.2

Catch and release

The practice of catch and release (also termed hook and release or live release) in rod fisheries has become increasingly common as a salmon management/conservation measure in light of the widespread decline in salmon abundance in the North Atlantic. In some areas of Canada and USA, catch and release has been practiced since 1984, and in more recent years it has also been widely used in many NEAC countries both as a result of statutory regulation and through voluntary practice.

The nominal catches presented in Section 2.1.1 comprise fish which have been caught and retained and do not include salmon that have been caught and released. Table 2.1.2.1 presents catch-and-release information from 1991 to 2004 for six countries that have records; catch- and-release may also be practiced in other countries while not being formally recorded. There are large differences in the percentage of the total rod catch that is released: in 2004 this ranged from 16% in Iceland to 76% in Russia, reflecting varying management practices among these countries. Within countries, the percentage of fish released has tended to increase over time. Overall, almost 144 000 salmon were reported to have been released around the North Atlantic in 2004, an increase of 12% on 2003, and the highest in the time series. There is also evidence from some countries that larger MSW fish are released in higher proportions than smaller fish.

Concerns have been expressed about the survival of fish following catch and release. Further details are included in Section 2.4.2.

2.1.3

Unreported catches

Unreported catches by year (1987–2004) and Commission Area are presented in Table 2.1.3.1.

A description of the methods used to evaluate the unreported catches was provided in ICES (2000) and updated for the NEAC Region in ICES (2002). In practice, the estimation methods used by each country have remained relatively unchanged and thus comparisons over time may be appropriate. However, the estimation procedures vary markedly between countries.

For example, some countries include only illegally caught fish in the unreported catch, while other countries include estimates of unreported catch by legal gear as well as illegal catches in their estimates. In France, nominal catches include a correction for under-reporting in rod fisheries. Over recent years efforts have been made to reduce the level of unreported catch in a number of countries (e.g. through improved reporting procedures). The introduction of carcass tagging programmes in Ireland and UK (N. Ireland) in recent years is also expected to lead to reductions in unreported catches.

The total unreported catch in NASCO areas in 2004 was estimated to be 686 t, a fall of 19%

from 2003 (847 t). The unreported catch in the North East Atlantic Commission Area in 2004 was estimated at 575 t, that for the North American Commission Area 101 t, with 10 t estimated for the West Greenland Commission Area. The unreported catch, expressed as a percentage of the total North Atlantic catch (nominal and unreported), has fluctuated since 1987 (range 23–34%; 25% in 2004), but has declined over the past 6 years (Figure 2.1.3.1). Esti- mates for 2004 are presented by country in Table 2.1.3.2. Expressed as a percentage of the total catch for the North Atlantic, these range from 0 to 12% for individual countries. Relative to national catches, unreported catches range between 1% and 57% of country totals.

(13)

In the past, salmon fishing by non-contracting parties is known to have taken place in international waters to the north of the Faroe Islands. Sixteen surveillance flights were made over the area in 2004. No sightings of vessels were made during the flights, although there was a period of five months over the winter when no surveillance took place. This is the period when salmon fishing has previously been reported.

2.2

Farming and sea ranching of Atlantic salmon

2.2.1

Production of farmed Atlantic salmon

The provisional estimate of farmed Atlantic salmon production in the North Atlantic area for 2004 is 796 839 t. This represents a small decrease on 2003 (803 488 t), but a 13% increase on the 5-year mean (1999–2003) (Table 2.2.1.1 and Figure 2.2.1.1). Most of the North Atlantic production took place in Norway (64%) and UK (Scotland) (22%).

World-wide production of farmed Atlantic salmon has been in excess of one million tonnes since 2002. Total production in 2004 is similar to that in 2003 and is provisionally estimated at a little over 1.1 million tonnes (Table 2.2.1.1 and Figure 2.2.1.1). Production outside the North Atlantic is currently estimated to account for about 30% of the total farmed production, with Chile (261 000 t) contributing the largest proportion of the production in this area.

World-wide production of farmed Atlantic salmon in 2004 was almost 550 times the reported nominal catch of Atlantic salmon in the North Atlantic. Farmed salmon therefore dominate world markets.

2.2.2

Production of ranched Atlantic salmon

Ranching has been defined as the production of salmon through smolt releases with the intent of harvesting the total population that returns to freshwater (harvesting can include fish col- lected for broodstock) (ICES, 1994). The total production of ranched Atlantic salmon in countries bordering the North Atlantic in 2004 was 12 t, an increase of 5 t on 2003 (Figure 2.2.2.1).

Salmon ranching (smolt releases) ceased in Iceland in 1998. Small catches of ranched fish were recorded in each of the three other countries reporting such fish (Ireland, UK(N. Ireland), and Norway), the data including catches in net, trap, and rod fisheries. Ranched fish comprised less than 3% of the nominal catches in each of these countries.

2.3

Update on the estimation of natural mortality at sea of Atlantic salmon

The Working Group was asked to examine further the evidence of changes in M for NAC and NEAC Atlantic salmon stocks, particularly in the second year at sea. The reviews of natural mortality undertaken by the Working Group in 2002 to 2004 were motivated by concerns that the value for natural mortality (M) assumed in the run reconstruction models was not appropriate (ICES, 2002, 2003, 2004a). In 2005, the Working Group reviewed further data from NAC and NEAC stocks to examine for trends in M over time and among stocks.

In 2002, the Working Group reviewed theoretical and empirical methods for estimating M for Atlantic salmon. After analysis of data from selected rivers, the instantaneous monthly mortality rate used in the run-reconstruction model for the North American and NEAC areas was changed from 0.01 to 0.03. In 2003, the Working Group reviewed an analysis of a data set from 5 rivers of the NEAC area extending from the Scorff (France) to the North Esk (Scot- land) and east to the Vesturdalsa River (Iceland) and 6 rivers in the NAC area, hatchery and wild stocks, extending from the Scotia-Fundy region to the north shore of the St. Lawrence (Quebec) (ICES, 2003). Both inverse weight and maturity schedule methods were applied to the stocks with appropriate data. The analysis of the river-specific growth data supported the previous conclusion that a linear function characterized the observed weights at age in the

(14)

marine phase better than the exponential function. The additional analyses confirmed the previous conclusion that monthly mortality in the second year at sea was greater than 1% and distributed around 3%, at least for the wild fish. There were important differences among stocks and even regions which were not accounted for in the generalization over the entire NEAC and NAC areas. It was previously noted that the estimates of mortality derived using the data sets did not account for all the fisheries removals, particularly those in the high seas, and were in fact estimates of Z (i.e. total mortality). Reductions in the level of fisheries should however have resulted in the fishing mortality component being a minor portion of the total mortality being estimated.

In 2004, the Working Group examined further the data requirements and assumptions of the models for estimating mortality at sea (ICES, 2004a). The analysis of series of return rate data from several rivers in both NEAC and NAC suggested that M could be higher than 0.03 in the last decade and in several stocks was increasing. Monitored return rates to many stocks in the NAC and NEAC areas are lower now under reduced exploitation than in the 1970s and 1980s when fisheries were more intensive suggesting that natural mortality must have increased as fishing mortality rate declined (Section 3.9.15, Section 4.9.9). The absence of sufficient data to model a temporally varying M in the run-reconstruction models of PFA adds to the uncertainty in the description of the recruitment and spawning stock functions. Large changes in mortality could however be detected under models with constant M as appears to be the case for the North American PFA dynamic. An analysis of changes in total mortality over time may provide an indication of the changes in exploitation if natural mortality is assumed to be constant over time.

2.3.1

An examination of temporal variation in mortality

In 2005, the Working Group reviewed return rates of hatchery and wild origin smolts to 1SW and 2SW from the NAC and NEAC areas. For NEAC, data were available for two hatchery origin stocks and six wild stocks with the longest temporal coverage for the wild and hatchery origin stock of the Burrishoole River (Ireland). For NAC, data from two hatchery origin stocks and five wild stocks were examined with the longest temporal coverage of 1970–2002 period.

Estimates of mortality could not be obtained from all year and stock combinations as some years had incomplete information for some of the age groups.

There was no evidence in the analyses presented that marine mortality (Z) had declined de- spite numerous fisheries reductions and closures, and in some stocks from both NAC and NEAC areas, there was an indication that mortality had increased (Fig. 2.3.1). The estimates from the maturity schedule model were consistently higher than from the inverse weight models. In the northern and southern NEAC areas, both wild and hatchery smolts show a constant decline in marine survival over the last two decades with the sharpest decline in the wild smolts of the southern NEAC area (Section 3.9.15). Similar declines in return rates of hatchery and wild salmon to the NAC area were also reported and returns rates of recent years were low compared to historic levels (Section 4.9.9). The Working Group reviewed evidence from the Miramichi River (NAC Canada) and the Teno River (NEAC Finland) that return rates of maiden salmon to repeat spawning has increased suggesting that the mortality factors affecting smolts in the first year and non-maturing salmon in the second year may be different from those interacting with previous spawners (Section 2.5).

The Working Group supported the extension of the smolt and adult monitoring programs for wild stocks of MSW salmon in the NAC area since 1996 and was encouraged that such activi- ties may provide additional data sets with which to track marine mortality. There continues to be an absence of monitored stocks in the northern areas of both NAC and NEAC. However, opportunities exist to monitor return rates from a large number of hatchery stocks (Table 3.9.15.2) and to assess their value as indicators of wild salmon survival at sea.

(15)

2.3.2

The use of hatchery reared salmon as an indicator of survival of wild salmon

Most of the index rivers where data on survival of wild smolts are available are small and mainly support 1 SW fish. The stocks from the index rivers are not representative for MSW stocks originating in large rivers. Because MSW stocks account for a significant proportion of the smolt production, it is important to estimate marine survival of salmon from these rivers too. However, tagging wild smolts in large rivers are very labour demanding and expensive, and it is thus unlikely that estimates of marine survival of wild salmon from such stocks will improve significantly in the near future.

In several countries large numbers of hatchery reared salmon are released for mitigation and experimentation. From numerous tagging experiments it has been shown that hatchery reared salmon are harvested in distant and home water fisheries side by side with wild salmon. There is, however, growing evidence that hatchery reared salmon may differ in life history and be- haviour from wild salmon. Hatchery reared salmon home to the river of release with less pre- cision than that of wild fish (Hansen et al., 1993; Jonsson et al., 2003).

Results from analyses carried out by SALMODEL (Crozier et al., 2003) show significant cor- relations between survival of wild and hatchery reared salmon from the same stocks and smolt cohorts. Additional information presented to the Working Group from Norway supported this and suggests that trends in survival of hatchery reared salmon might be used as an index of survival of wild salmon. In recent years there have been improvements in smolt rearing which may have resulted in enhanced parr growth and subsequent increase in smolt size. This may have contributed to the observations from Norway in recent years that show survival of hatchery reared smolts was similar to that of wild smolts, whereas in the 1980s survival of wild smolts tended to be superiour to that of hatchery reared smolts.

The provisional results from the analyses of survival of wild and hatchery reared smolts show a potential for using hatchery reared smolts as a tool for estimation of survival and survival trends of wild salmon. However, the survival time series should be analysed in more detail particularly to adjust for changes in smolt quality of hatchery reared salmon. Furthermore, there are data available from Ireland, UK (Northern Ireland) and Iceland that could be included and refined. The success of a hatchery programme in this respect is highly dependent on a satisfactory and stable smolt quality over time, and a representative genetic structure for the stock to be released. It is important to develop a smolt production strategy that makes the fish fit for survival in nature and more representative of wild smolts. This may require a different rearing scheme than that to produce smolts for farming. Furthermore, to obtain the best return rates, a tag and release strategy should be developed as well as a strategy to recover tags.

2.4

Significant developments towards the management of salmon

2.4.1

Progress in developing precautionary catch advice for Irish salmon fisheries

The Precautionary Approach to salmon fisheries management adopted by NASCO (NASCO, 1998) includes the following:

• that account be taken at each stage of the risks of not achieving the fisheries management objectives by considering uncertainty in the current state of the stocks, in biological reference points and fishery management capabilities, the formulation of pre-agreed management actions in the form of procedures to be applied over a range of stock conditions.

(16)

The methodology for setting conservation limits (CL) for Irish salmon stocks was changed in 2003 from a stock and recruitment curve based on district catch data, unreported catch and exploitation rate data (Potter et al., 1998, Ó Maoiléidigh et al., 2001) to a Wetted Area/

Bayesian hierarchical stock and recruitment model, BHSRA (Prévost et al., 2003, Ó Maoilé- idigh et al., 2004).

In 2004 and 2005, the catch advice for Irish fisheries was modified to take account of the risk of not achieving fisheries management objectives (conservation limits in all rivers within a district), uncertainty in biological reference points (i.e. sex ratio and required egg deposition) and the formulation of pre-agreed management actions in the form of procedures to be applied over a range of stock conditions (harvest guidelines).

Risk analysis and derivation of precautionary catch advice

The methodologies used to modify the wetted area conservation limits are described in Crozier et al. (2003). Once estimates of average spawners, average catch, and district CL are produced, precautionary catch advice is formulated by applying harvest guidelines in a risk analysis. ICES (2003) recommended the adoption of a 75% probability level of meeting conservation limits for salmon stocks. This recommendation was also taken into consideration in the provision of catch advice.

Increasing the probability of achieving the required number of females Although the egg numbers are converted to fish for advice purposes using average biological characteristics, ultimately, the objective is to ensure with high probability that the required number of eggs is achieved, or in a simpler case, that the number of female salmon required is achieved. As such, it is necessary to know how many fish must return to a river to spawn (i.e.

escape the fisheries) to ensure that there is a sufficient number of females. To ensure (i.e. with high probability) that the required number of females will be achieved, the total (males and females) required to escape the fisheries must be calculated (males and females must be considered together, because the fisheries cannot selectively harvest one sex over the other).

For example, the probability of obtaining at least 100 females if the returning stock has a 50:50 female to male ratio and exactly 200 fish are released from the fishery is only 50%. This would not be an acceptable level of risk under the Precautionary Approach. Therefore the number of females which would need to escape each fishery to provide a 75% chance of achieving the correct egg deposition has been calculated. In most districts, this is a minor adjustment (moving from 50% to 75%) because the CL for fish is in the thousands. As an example, for the Waterford district, this correction amounts to less than 2% more fish.

Increasing the probability that the CL will be achieved in every river in the district at the same time

Since river-specific CLs have been defined, there is a further objective in ensuring that each individual river meets its CL at a high probability level. Although, the district CLs are based on the individual CLs summed, this number of fish will not be sufficient to ensure that each river meets it’s CL as the distribution of the fish will not necessarily match the proportions required for each individual river. A fish released from the mixed stock fishery may be female from river 1, male from river 1, female from river 2, male from river 2, female from river k, male from river k (where k represents the total number of rivers in the district). It is possible to calculate the number of fish which will meet the CLs in each river simultaneously at a high probability (i.e. 75 percent of the time). The end result is that for example, in the Waterford district CL must be increased by 18% in order to have a 75% chance that the river specific CLs will be achieved simultaneously in the 14 rivers in this district.

(17)

Risk analysis for the catch advice

Previously, using the deterministic calculation (i.e. using the average returns for the period 2000–2004 as the forecast for 2005) and calculating the surplus (returns – CL), there was a 50% chance of meeting or exceeding the CL (or conversely, a 50% chance of not meeting it).

However, it is desirable to know the harvest level which would provide a 75% chance (as recommended by ICES) of meeting or exceeding the CL. Initially, a random series of returns is drawn from a normal distribution of the estimated returns for the period 2000–2004. A se- quence of harvests (from 0 upwards) is subtracted from these returns to provide the estimated number of fish remaining for spawning. The number of times that the spawners equal or exceed the CL is counted for each level of harvest. This count is used to generate a risk plot which provides the probability of meeting CL at varying levels of harvest i.e. the higher the level of harvest on the returning stock the less likely are the chances of meeting the CL. The harvest level corresponding to a desired probability of meeting the CL: 0.75, 0.9, 1.0 etc. can be read off the plot.

Harvest guidelines and precautionary catch advice

The following harvest guidelines apply to the precautionary catch advice:

• Generally, the harvest option providing a 75% chance of meeting the CL in a given district is chosen as the precautionary catch advice (Figure 2.4.1.1).

• In following a precautionary approach, increases over the average catch for the period 2000 to 2004 should not be permitted even if the harvest option at the 75%

probability of meeting the CL is higher. This is because each district fishery catches salmon destined for other districts and there is clearly a need to protect vulnerable stocks in these other districts. This advice will be reviewed annually to assess any improvement in the status of these vulnerable stocks (Figure 2.4.1.2).

• Where there is no harvest option which will provide a 75% chance of meeting the district CL, then the precautionary catch advice is that there is no surplus of fish to support a harvest (commercial or rod). This is illustrated in Figure 2.4.1.3.

This advice is predicated on wild fish only (i.e. estimated returns from hatchery-released smolts have been removed). It also relates to the total removal of fish by all means, and is not restricted to commercial fisheries. There are eight districts, mainly located on the east and south coasts, where the CL will probably not be met even in the absence of harvests of salmon. In six of the districts, reductions in the average catch in 2005 are indicated if there is to be a 75% chance of meeting the CL. The remaining districts are meeting or exceeding their CLs. In this instance, the average catch is advised for 2005, even where the harvest option providing a 75% chance of meeting the CL is higher. This recognizes the fact that these fisheries intercept salmon destined for districts that are below their CL. The status of these districts will be assessed on an ongoing basis, and the advice will change in line with any significant and consistent improvement in stock size. The maximum harvest by all methods being recommended is 122 541 one-sea winter salmon for the 2005 season.

The scientific process continues to be developed and improved in line with best international practice (including the precautionary approach as adopted by NASCO) and other areas which warrant further attention are:

• the mixed stock nature of district fisheries and the difficulties in assigning stocks in catches to district of origin.

• the accuracy of the wetted area estimate, which is based on a regression model with associated uncertainty.

• deterioration in water quality is a serious issue in a number of districts in east and southeast Ireland (i.e. Dundalk, Dublin, Wexford, and Waterford) and will influ- ence the productivity of the rivers in meeting CLs.

(18)

• the models used for district catch advice do not include the 2SW and MSW (or repeat spawner) fish in the CL (i.e. no input data for this component is entered in the model). In some areas this may underestimate the available spawner capacity.

• information from existing automatic fish counters which may be used to manage on a specific river basis or be used as an index of several or all rivers in a district.

2.4.2

Catch and release

The practice of catch and release in salmon rod fisheries has become a common management/conservation measure. In some areas of Canada and USA, anglers have been required to practice catch and release since 1984. More recently it has also been widely used in many NEAC countries both as a regulation and a voluntary practice. The percentage of the total rod catch released in 2004 was 55% in Canada and ranged from 16% in Iceland to 75% in Russia (Table 2.1.2.1). Within countries, the percentage of salmon released by anglers has tended to increase over time. In 2004, anglers reported releasing almost 144 000 salmon around the North Atlantic, the highest number in the time series. Larger MSW fish are released in higher proportions than smaller fish in some countries.

In assessing the attainment of river-specific conservation limits, Canada (various regions) and UK (England & Wales) make a small allowance for catch-and-release mortality. These correc- tions vary; up to10% for Canadian regions and 20% for UK (England & Wales). The allow- ances are based on research studies (detailed in ICES, 2003) documenting that fish are more likely to die when water temperatures are high (>20^oC) or if fish are ‘played’ for an extended period.

The Working Group reviewed a probabilistic method to predict the risk of mortality for caught and released salmon from the Penobscot River, USA. Periods of similar thermal conditions based on median daily temperatures in the catch and release fishery area were identified. For each period, hooking mortality for the number of salmon reported caught and released was drawn from a uniform probability distribution based on the literature (i.e. 0.05 to 0.30 for temperatures greater than 20^oC). Ten thousand simulations drawing from a random binomial function were done to calculate the number of hooked and released fish that died. The resulting distribution describes the mortality losses from reported numbers of caught and released salmon. Hooking mortality was assumed to be uniform for all salmon (origin, sea-age, and sex). These simulations suggest that mortality following capture can be low. A recent radio tracking study in northwest England found that upwards of 85% of released spring salmon can be expected to survive to spawning (UK Environment Agency, 2003).

The survivors of catch and release angling are vulnerable to being hooked again. Additional information from rod fisheries on four Icelandic rivers documented that 24.4% (range: 22.1–

27.8%) of salmon were captured for a second time. Salmon captured a third time were rare (1.8%). Exploitation rates in these rivers range from 45% to 60%. These results provide a means for adjusting the catch and release statistics to account for multiple recapture in these rivers and potentially for Iceland as a whole.

The Working Group encourages studies and further development of probabilistic methods that assess catch and release angling mortality. Models that can relate angling mortality to varying levels of effort would be especially useful. They also noted the need for further studies on the rate of multiple recaptures in catch and release angling in other rivers, especially those where the total salmon run is known.

2.4.3

Regional growth patterns

Pre Fishery Abundance (PFA) of Norwegian salmon has been developed at the national level.

It is of interest to estimate PFA for smaller units, as marine survival may vary among river stocks and regions. Therefore it may help to identify survival signals other than return rates of

(19)

adults from smolt cohorts. One such signal that could be applied for grouping stocks is growth. It has been shown that survival and growth of salmon at sea during the first year are correlated (Friedland et al., 2000).

Systematic collection of salmon scales from anglers’ catches has been carried out in seven rivers in Norway. Back-calculated growth of the first year at sea of 1SW fish was systemati- cally lower than that of MSW fish of the same smolt year class in all rivers. For six of the seven river stocks, the first year growth of 1SW and MSW fish of the same smolt year class were significantly correlated (P=0.000–0.018).

Growth of salmon the first year at sea varied among years and stocks, with a systematic trend for slower growth in salmon originating from Northern latitudes. There were significant corre- lations in growth between salmon originating from nearby rivers (P=0.000–0.024), whereas growth in more distant stocks were less correlated (P=0.007–0.44).

The marine growth of the four most northerly salmon stocks was significantly correlated with the mean sea temperature at 50 m depth in the Norwegian Sea (66° N; 2°E) and mean temperature in the 0–50 m layer in the Barents Sea (70°30’ N – 72°30’ N; 33° 30’ E) during July- December. However, the most northerly populations were more strongly correlated with temperatures in the north than with temperatures in the southern area. Growth of salmon from rivers in mid Norway showed highest correlation with temperatures in the Norwegian Sea.

Salmon growth from the three most southern rivers was not correlated with temperatures at any of the two stations.

The present results indicate that PFA of Norwegian salmon should be developed at regional levels to improve the estimates. At least two regions should be established, divided by the Lofoten Islands at a latitude of 68° N. The main mortality of salmon seems to take place in the early post-smolt phase. Therefore, ocean climate during this phase is expected to be important for growth and survival. Ocean climate varies considerably along the coast of Norway, and hence, salmon from different stocks experience considerable differences in climate when they migrate to sea.

Further support to the regional grouping of rivers is provided by analyses from three subarctic rivers running to the Barents Sea within a small geographic area in northeastern Europe.

Salmon from the rivers Teno/Tana (Finland &Norway), Näätämöjoki/Neidenelva (Finland &

Norway), and Kola (Russia) showed significant temporal synchrony in marine growth and variation in abundance, and these variables were also significantly correlated with the sea water temperature in the Barents Sea.

2.5

Long-term projections for stock rebuilding

In 2003 and 2004, the Working Group provided information on long term trajectories for stock rebuilding for specific stocks with different productive capacities and under different conditions of exploitation and starting stock size (relative to CL). The data and analysis indicated that there is an increased probability of not achieving Slim in low productivity rivers when exploitation was increased and that stock productivity was the most important factor in determining the ability of a stock to rebuild in a mixed stock fishery.

The Working Group therefore cautioned that further simulations should also reflect declining stock trajectories and population viability given that the probability of rebuilding in the short term is low in most areas and that the main result of recent management measures may have been to reduce this rate of decline rather than lead to any significant stock rebuilding.

(20)

2.5.1

Long term projections of PFA for North America

Seven different types of regression models have been used to relate lagged spawners (LS and Pre-Fishery Abundance for North America (PFANA, Section 5.10). Some of these allow for a

“regime shift” in the relationship identified by the Working Group (ICES, 2003) whereby early years in the time-series demonstrate higher PFA per lagged spawner while the more recent years demonstrate lower PFA per lagged spawner. The LS value for the current year is used to predict a distribution of expected PFANA in the current year which is used to provide catch advice for the upcoming fishing season in West Greenland. However, medium term (up to 5 years) and long term (up to 20 years) projections have not been developed to date. There- fore the Working Group has adapted and extended the analysis in Section 5.10 to complete the cycle over a longer time period to examine potential long term trajectories in stock size. The only new assumption made is that the allocation by region of surviving fish after the West Greenland fishery in year t is proportional to the distribution of the lagged spawners by region that produced the predicted PFANA for that year. This additional assumption allows medium and long term predictions for PFANA to be made, demonstrating directly the implications of the different relationships between LS and PFANA and also providing a basis for comparison with the simple Population Viability Analysis (PVA) results presented to the Working Group in (ICES, 2004a).

The basic data required for reconstruction of the historical PFANA are given in Section 5.10.

The wide range of functional relationships between LS and PFANA used in the model, along with the selection procedure to pick the best model, are described in Section 5.8. The predictions in Section 5.8 utilize the measured uncertainty in the regression fit to make stochastic projections of PFANA for the upcoming year. This has been the end point of this process to date. However, the cycle can be completed by assuming that the PFANA that is predicted for next year can be allocated to the six regions based on their LS relative abundance (Figure 2.5.1.1). This is a reasonable assumption because the PFANA value is formed from the returns, which are highly correlated to spawners (perfectly so, if in-river catches are subtracted from the returns). Once the PFANA is reduced by the West Greenland fishery, allocated to the Sea Fishing Areas (SFAs), and these values reduced by natural mortality and any home water fisheries, the spawners for next year can be derived. Application of the lagging process can then be used to estimate future values of LS and this cycle can be repeated indefinitely into the future.

When stochastic projections are used in the Cycle approach, then uncertainty of PFANA for every year into the future can be derived by incorporating the appropriate root mean square error from that simulation’s model fit. These distributions can be used to show the probability of stock increase or decrease given the assumed natural and fishing mortality rates and the LS- PFANA relationship chosen. Due to both the large amount of uncertainty in the LS-PFANA relationship, as well as the use of power functions for this relationship, the stochastic projections can sometimes produce exceedingly large values for PFANA. To prevent numerical overflows, a cap is placed on the maximum possible value for PFANA in any year. The cap was arbitrarily set at five million fish, approximately five times greater than the largest PFANA estimate in the observed time series.

Conducting stochastic projections using this approach under the assumption of no fishing results in a wide range of possible PFA values in the next 20 years, with major differences seen between projections made in the high phase versus the low phase (Figure 2.5.1.2). The high phase projections are limited by the ad hoc cap of 5 million fish. Separating the projections by model type reveals that the model selected for a particular simulated dataset determines the overall result of medium to long term projections, as expected (Figure 2.5.1.3).

(21)

• This is most clearly seen by comparing model 3 (constant PFA independent of lagged spawner size) with models 4 through 7 (PFA depends upon lagged spawners).

• Models 1 and 2, which do not have a regime shift, were never selected (Table 2.5.1.1).

• In model 3, the projections randomly fluctuate about a mean because no matter how large or small the number of lagged spawners gets, the expected value of PFA is always the same. The median PFA value for the model 3 high phase projections is approximately 450 000, much higher than the median from the model 3 low phase projections.

• The median PFA for models 4–7 always increase to the ad hoc cap when pro- jected in the high phase. Of these model projections, only model 7 has some probability of not increasing to the ad hoc cap of 5 million fish in the next 20 years.

• In contrast, none of the projections of PFA using models 4–7 in the low phase are limited by the ad hoc cap of 5 million fish. The projections of model 4 in the low phase increase slightly over the next 20 years while the projections of models 5-7 in the low phase decrease. The decrease is most severe for model 7, with median PFA declining to less than one fish by 2022.

• The projections of models 5–7 in the low phase and the simple PVA presented in ICES (2004a) are similar.

• Eliminating projections under the highly unlikely assumption for long-term projections that LS and PFA are unrelated (models 1 and 3), results in PFA values that, under a scenario where there is no fishing, the PFA is expected to either increase rapidly for the high phase projections or slightly decrease for the low phase projections (Figure 2.5.1.4).

Further PFA projections were next made assuming that a fishery occurs in West Greenland at 20, 50, or 100 t annually for the next 20 years, and that all home river fisheries are stopped (Figures 2.5.1.5-7). The harvest in tons was converted to numbers of fish and split between North America and Europe following the standard WG approach. In all three West Greenland harvests considered, PFA from the high phase projections were essentially the same with large increases up to the ad hoc cap. In contrast, PFA from the low phase projections showed a strong response to West Greenland harvest, with a continued harvest of 100 t causing the median PFA to decline to zero by 2013 (Figure 2.5.1.8). The overlap of medians for the first five years of projections is due to the fact that the lagged spawners that produced these PFA values come from spawners that are already back in the river for most SFAs. Thus, there is no feedback from the cycling nor from forecast catches in West Greenland in this period.

Models that did not relate PFA to LS (models 1 and 3) were selected 15% of the time. How- ever, these are not included in Figures 2.5.1.4-8 as they are not informative for medium and long term projections.

These results demonstrate that medium to long term forecasts of PFA depend most on the phase used for projections. The PFA is much more resilient to fishing when in the high phase than when in the low phase. The ability to detect a switch from the current low phase to the high phase depends on future PFA estimates from observed returns that are much higher than expected from the low phase model. A single observation is not sufficient to claim that a change in phase has occurred, multiple years in the high phase will be required. There is a time lag between observing large PFA and the feedback through the cycle to generate higher returns, spawners, and PFA that needs to be considered when making management decisions.

Future development of long term trajectories should consider the following issues.

• Inclusion of ages other than two, i.e. grilse and multi-sea winter ages, should be considered to more accurately reflect the biology of Atlantic salmon.

(22)

• The lagged spawner calculations assume equal contribution to PFANA from the six SFAs. This assumption should be examined to determine if weighting of the SFA would produce a more biologically realistic index.

• The development of juvenile indices could improve the models’ ability to forecast PFANA by including another life stage. Adding another audit point to the cycle would either confirm and stabilize the cycle or else demonstrate that it is in- correct, which could be due to any number of reasons such as differences among region of contribution to PFA, differing impact of high sea or coastal fisheries on the regions, or different natural mortality rates experienced in the different regions.

2.5.2

Potential for rebuilding two multi-sea-winter salmon stocks of the maritime provinces

The ICES (ICES, 2003) catch advice for the management of the West Greenland fishery and the management of the homewater fisheries has been provided on the basis of achieving the conservation objectives of the four northern regions (Labrador, Newfoundland, Quebec, and Gulf) and to an alternate objective for the southern regions of achieving at least a 10% increase or a 25% increase relative to the average returns to the regions during a specified time period (Chaput et al., 2005). In this regard it is presumed that stocks in these areas have the potential to rebuild if adequate spawning occurred.

The stock status and short term viability and potential for rebuilding of two multi-sea-winter salmon stocks (Mirimichi River and Saint Johns River) of the maritime provinces of eastern Canada are examined in the context of the ICES advice. Age data were used to assign salmon to their year class (year of egg deposition or year of adult return to the river). Mean fork lengths and proportion female were estimated for the origin, size group, sea age, spawning history and fresh water age combinations in any given year. The eggs per female were calculated from river-specific egg to length fecundity relationships. Eggs in the returns and spawners by origin, size group, sea age, spawning history and fresh water age were calculated from the estimated returns by origin (hatchery and wild) and size group, the proportion female in the age group, and the average eggs per female fish of the age group.

Age structure and relative abundance

Characteristics of these two populations were examined in detail in 2004 (ICES, 2004a). The Atlantic salmon population in the Miramichi is characterized by an expanding spawning age history structure (Table 2.5.2.1). Between 1971 and 1986, there were few repeat spawners in the river with at most two previous spawning migrations. Since 1992 and 1995, adult salmon on their sixth and seventh spawning migrations, respectively, have been sampled in the catches at the estuary trap nets and repeat spawning salmon have comprised 6% to 21% of the total returns of all age groups.

In the Saint John River, repeat spawners made up as much as 7% of the total returns to the river but this was in the early part of the time series, 1970s, and in the last ten years, repeat spawners have represented less than 3% of the total returns with a virtual absence of any salmon beyond a second spawning (Table 2.5.2.2).

As the repeat spawner abundance has increased in the Miramichi, the number of year classes present in the annual spawning migration has increased from four to five in the 1970s to as many as eight to nine year classes in the returns of the 1990s. This contrasts with the Saint John River stock where for either wild salmon or hatchery origin salmon, there has been no change in the number of year classes over time, with generally five year classes present, with a maximum of seven in the earlier portion of the time series.

(23)

Replacement ratios of egg production

For a population to replace itself, one egg in the recruitment is required for every egg spawned. For the Miramichi River, the wild salmon stock produced maiden fish recruitment surplus to spawners for most year classes between 1971 and 1989 but was consistently below replacement for the 1990 to 1997 year classes (Figure 2.5.2.1). The eggs in the maiden returns of the 1998 year class (the last year where an assessment was possible) are estimated to have been equivalent to the eggs which were spawned.

For the Saint John River, wild salmon production has varied around the replacement line for the 1972 to 1988 year classes but has decreased sharply and remains well below replacement for the 1989 to 1999 year classes (Figures 2.5.2.2, 2.5.2.3). There are so few repeat spawners in this stock that the difference in the lifetime production of eggs relative to production from maiden salmon is minimal. An examination of the spawners to recruitment plot indicates that the population was sliding down along the replacement line but egg to maiden salmon mortality increased substantially and has remained high for the 1989 and subsequent year classes (Figure 2.5.2.2). Replacement ratio calculations for the hatchery origin salmon indicates a relatively constant rate of return for the 1985 to 1999 year classes (Figure 2.5.2.3).

There has been a decline in the proportion of the eggs produced in the lifetime of the year class by maiden MSW salmon of the Miramichi River (Figure 2.5.2.4). For the 1981 and subsequent year classes, the lifetime egg production from MSW maiden salmon amounts to about 50% of the lifetime production of the year class. The decline from the previous time period is attributable to a decline in 2SW maiden salmon abundance and an increase in the repeat spawner abundance. This contrasts with the Saint John River stock in which the MSW maiden salmon continue to contribute over 80% of the eggs in the lifetime of the year class with a slight decline for the recent three year classes (Figure 2.5.2.4). In the hatchery origin salmon, 1SW maiden salmon can contribute as much as 25% of the eggs during the life time of the cohort and this has increased for the post 1989 year classes.

Constraints to rebuilding in these Maritimes rivers

The replacement ratio calculations indicate that the Miramichi River population had the potential historically to produce a surplus of maiden egg production. In the recent decade, maiden egg production (i.e. recruit egg production) has been well below replacement for a time period when spawner egg depositions exceeded the conservation requirements by 50 to 100%. With a decline in egg depositions, returns of maiden fish have resulted in at least the replacement of the parental stock and the potential for increased returns appears feasible. The decline in 1SW and 2SW maiden salmon abundance may have been partly the result of high egg depositions for those year classes and resulting over compensatory density dependent response in freshwater. Freshwater conditions have remained suitable for the survival of high numbers of juveniles from fry to age-1 parr and from age-1 parr to age-2 parr (Figure 2.5.2.5). Since 1999, smolt production from the Northwest Miramichi, one of the two main branches, has ranged from 162 to 390 thousand, a production rate of 1.0 to 2.3 smolts per 100 m² of rearing habitat area (Chaput et al. 2002). This is less than expected given one-year old and two-year old parr densities ranging between 30 and 47 fish per 100 m² (Figure 2.5.2.5). The total biomass of juveniles at the monitored sites in the Northwest Miramichi has averaged 230 to 377 grams per 100 m² in the years prior to the smolt emigration year. There is a negative association between the mean biomass of juveniles of the year and the smolt production the following spring. The true dynamic is not well defined because there are no observations for low to medium biomass levels for the river but parr age-1 index plotted against the returns of maiden 1SW and 2SW salmon adjusted to the year of smolt emigration supports the view of a strong over-compensatory relationship in salmon production for the Miramichi. An examination of the spawners to recruitment plot from Figure 2.5.2.2 indicates that there is a greater chance (9 of 11 events) that the recruitment will be less than the spawners when egg depositions exceed

Main Report (5.722Mb)

ICES Advisory Committee on Fishery Management ICES CM 2005/ACFM:17 R

. I

Report of the Working Group on North Atlantic Salmon (WGNAS)

5-14 April 2005

Nuuk, Greenland

Conseil International pour l’Exploration de la Mer

Contents

1 INTRODUCTION

2 ATLANTIC SALMON IN THE NORTH ATLANTIC AREA

Catches of North Atlantic salmon

Nominal catches of salmon

Catch and release

Unreported catches

Farming and sea ranching of Atlantic salmon

Production of farmed Atlantic salmon

Production of ranched Atlantic salmon

Update on the estimation of natural mortality at sea of Atlantic salmon

An examination of temporal variation in mortality

The use of hatchery reared salmon as an indicator of survival of wild salmon

Significant developments towards the management of salmon

Progress in developing precautionary catch advice for Irish salmon fisheries

Catch and release

Regional growth patterns

Long-term projections for stock rebuilding

Long term projections of PFA for North America

Potential for rebuilding two multi-sea-winter salmon stocks of the maritime provinces